HERPETOLOGICAL JOURNAL 21: 219–225, 2011 Reproduction of Xenodon dorbignyi on the north coast of Rio Grande do Sul, Brazil Roberto Baptista de Oliveira1,2, Gláucia Maria Funk Pontes1,2, Ana Paula Maciel1, Leandro Ribeiro Gomes1,2 & Marcos Di-Bernardo1,2* 1Laboratório de Herpetologia, Museu de Ciências e Tecnologia, Pontifícia Universidade Católica do Rio Grande do Sul, Brazil 2Faculdade de Biociências, Pontifícia Universidade Católica do Rio Grande do Sul, Brazil *In memoriam Information on sexual maturity, reproductive cycle, fecundity and sexual dimorphism of Xenodon dorbignyi was obtained from 537 individuals captured in a 333 ha sand dune area on the north coast of Rio Grande do Sul, Brazil, and from dissection and analysis of gonads of 98 specimens from the same region deposited in scientific collections. Females and males reach sexual maturity at about 260 mm and 220 mm snout–vent length (SVL), respectively. Males reach sexual maturity in their first year, whereas some females mature in their second year. Both sexes reach similar SVL; males have a relatively longer tail, and mature females have a heavier body. Reproduction is seasonal, with vitellogenesis occurring from August to January, mating from August to December, oviposition from November to February, and hatching from January to April. Clutch size varied from three to 10 eggs and was correlated with maternal SVL. The ratio between clutch mass and female total mass varied between 0.19 and 0.42 (>0.30 in 80% of observations). Key words: Dipsadidae, natural history, reproductive cycle, Serpentes, sexual dimorphism, southern Brazil INTRODUCTION do Sul, gathered from a large number of observations in nature and from dissections of specimens deposited in he life history of an organism is characterized by re- collections. Tproductive mode, reproductive frequency, clutch and litter size, size of newborns and age and size at sexual ma- MATERIALS AND METHODS turity (Pough et al., 2004). As a consequence of different reproductive costs, the age and size at which individuals The northern coast of Rio Grande do Sul is mainly char- mature can vary both inter- and intraspecifically (Parker acterized by a plain of mobile dunes, interposed by small & Plummer, 1987), leading to sexual size dimorphism at humid depressions and patches with scarce vegetation, the adult stage (Shine, 1993). Fecundity is an important mostly Graminae (Waechter, 1985). The climate is tem- determinant for female reproductive success (Bonnet et perate, with mean, minimum and maximum monthly al., 2001; Lourdais et al., 2003), in addition to timing of temperatures ranging between 15.4, 12.2 and 18.3 ºC in oviposition and quality of offspring (e.g. Olsson & Shine, July, and 24.8, 21.3 and 27.6 ºC in February, respectively. 1997; Madsen & Shine, 1998). High reproductive out- Rainfall is almost uniform across the year, with a small put can be limited by environmental, genetic, behavioral, peak during winter and a mean of 1323 mm (Hasenack & physiological and morphological factors (Williams, 1966; Ferraro, 1989). Shine, 1992). Field observations were carried out from July 1998 Xenodon dorbignyi (Duméril, Bibron & Duméril, to June 2004 in a 333 ha area of sand dunes located in 1854) is an oviparous dipsadid snake (tribe Xenodonti- Magistério (30º21'S, 50º17'W), municipality of Balneário ni) distributed in Paraguay, Argentina, Uruguay and the Pinhal. southernmost parts of Brazil, where it occupies open en- Complementary information was gathered from 98 vironments (Gudynas, 1979; Lema, 1994; Tozetti et al., specimens (30 males and 68 females) collected locally 2010). Information on its reproductive biology mostly and deposited in the herpetological collection of the comprises records of clutches, hatchlings born in captivity Museu de Ciências e Tecnologia of Pontifícia Universi- and occasional observations of mating, mainly from pop- dade Católica do Rio Grande do Sul (MCP), Fundação ulations in Uruguay and Argentina (e.g. Orejas-Miranda, Zoobotânica do Rio Grande do Sul (MCN), and Depar- 1966; Gallardo, 1977; Gudynas, 1979; Leitão-de-Araújo, tamento de Zoologia of Universidade Federal do Rio 1978); Pontes & Di-Bernardo (1988) presented the only Grande do Sul (DZUFRGS) (Appendix 1). Specimens reproductive data available for populations from Rio from collections were dissected, recording snout–vent Grande do Sul. length (SVL), tail length (TL), sex, number of eggs or This study presents information about sexual maturity, vitellogenic follicles, size of the larger follicle or egg (in sexual dimorphism, reproductive cycle and fecundity mm), and condition of deferent ducts. Individuals were of Xenodon dorbignyi on the north coast of Rio Grande considered mature when they had vitellogenic follicles Correspondence: R.B. Oliveira, Laboratório de Herpetologia, Museu de Ciências e Tecnologia, Pontifícia Universidade Católica do Rio Grande do Sul, Av. Ipiranga 6681, 90619-900 Porto Alegre-RS, Brazil. E-mail: [email protected] 219 R.B. Oliveira et al. larger than 8 mm or eggs in oviducts (females), or when they had convoluted and opaque deferent ducts (males, cf. Shine, 1982). For ambiguous specimens, transverse slices of the median region of the testis were prepared on slides, stained with haematoxilin-eosin and analysed for the presence or absence of spermatozoa. Maturity of live individuals was inferred based on SVL. Sexual dimorphism was evaluated based on 537 live specimens (260 males and 277 females), and on 100 spec- imens (40 males and 60 females) born in captivity from gravid females captured in the study area. All individuals were measured (SVL and TL, using a ruler), weighed (in grams, using Pesola spring scales), and sexed by forced eversion of the hemipenes. The animals were individually marked by branding ventral scales with a burning wire (cf. Di-Bernardo et al., 2007), and released at their site of capture. After the branding process, marks were treated with antiseptic spray and an anaesthetic (benzalkonium chloride plus lidocaine) to avoid infection by pathogens. Only data from the first capture were used. In order to avoid the inclusion of possibly gravid females in the body mass analysis, only data from mature individuals collect- Fig. 1. Seasonal distribution of snout–vent length ed from March to July (out of the reproductive season) of females (above) and males (below) of Xenodon were included. Since the SVL of both mature males and dorbignyi. The stippled line represents the snout–vent females violated the assumptions of equality of variance, length of the smallest mature individual. an adjusted Student’s t-test was used (t') to compare their means (cf. Zar, 1999). Differences in TL and body mass were investigated by comparing their regressions in rela- tion to SVL. The homogeneity of slopes was compared through analyses of variance (ANOVA), and the inter- cepts were compared with each other using covariance analysis (ANCOVA) after ln transformation of data. The vitellogenic period was inferred from the seasonal RESULTS distribution of the larger follicles of dissected females. The mating season was determined from observations of Sexual maturity mating pairs. Gravid females were kept until oviposition SVL of females (n=67 preserved specimens) varied be- (1–15 days at 24 oC), before release at their capture sites. tween 178 mm and 469 mm; the smallest mature female These females were maintained individually in boxes of was 257 mm. SVL of males (n=31) varied between 160 mm polyethylene sterilized with iodine solution. Eggs were and 451 mm; the smallest mature male was 192 mm. One counted and weighed within 48 hours of oviposition, and of two immature males had an SVL of 207 mm, which incubated in humid, sterilized vermiculite at about 28 ºC. was larger than the SVL of two mature males. All males Hatchlings were counted, measured, weighed, marked and above 207 mm SVL were sexually mature. released at their mother’s capture sites after they hatched. The seasonal distribution of body size classes indicated Relative clutch mass (RCM) was calculated as the ratio that males attain sexual maturity in the first reproductive between clutch mass and female total mass (including the season after their birth (Fig. 1). The capture of small in- clutch mass) (cf. Seigel & Fitch, 1984). The relationships dividuals in December suggests that some females do between the size of eggs, hatchlings and females were not attain sexual maturity in the first reproductive season investigated using linear regression. (Fig. 1). Reproductive frequency was estimated from the pro- portion of gravid females found during the reproductive Sexual dimorphism season (cf. Plummer, 1984). Only captures from Novem- SVL of females (n=337) varied between 104 mm and ber were used because in October many females did not 490 mm, and SVL of males (n=300) varied between have well developed follicles that could be detected by 114 mm and 500 mm (Fig. 2). There was no signifi- palpation, and in December several females had already cant difference between the SVL of newborn females laid eggs. and males (n=60 ♀ and 40 ♂; x=134.9±10.9 mm and Distributions were tested for normality and homogene- 136.8±10.9 mm, respectively; t=0.85, df=98, P=0.397). ity of variances using Kolmogorov–Smirnov and Levene The mean SVL of mature females was significantly tests, respectively. All analyses used α=0.05. Mean val- higher than that of mature males (n=188 ♀ and 239 ♂; ues are presented with ±1 standard deviation. Statistical x =351.1±45.0 mm and 321.3±57.20 mm, respectively; analyses were performed with SPSS 10.0 for Windows t'=6.04, df=425, P<0.001), although individuals of both (SPSS Inc., 1999). sexes attain similar maximum SVL (Fig. 2). 220 Reproduction of Xenodon dorbignyi Fig. 2. Relationship between tail length and snout–vent Fig. 3. Relationship between body mass and snout– length of female (circles) and male (triangles) Xenodon vent length of female (circles) and male (triangles) dorbignyi.